A Simple Scheme of Channel Estimation in Large MIMO Systems

Abstract

We address the problem of channel estimation in a large multiple-input multiple-output (MIMO) system using pilot symbol aided algorithm in Rayleigh fading environment. Conventional minimum mean square error and minimum variance unbiased based channel estimators require the inversion of covariance matrices, which has computational complexity of the order of N^6, where N is the number of antennas at source and destination. When there are hundreds of antennas, the aforementioned approaches are extremely complex. Recently, low complexity decoders based on polynomial expansion are proposed with computational complexity of the order of N^4. We propose a very simple scheme of channel estimation in large MIMO system, in which we decompose the channel matrix in 2 × 2 matrices and then we estimate these 2 × 2 matrices in order to reduce the computational complexity to the order of N^2 and making the system for practical use. Further, a decoder of OSTBC data is derived by assuming the perfect channel state information (CSI) at the destination; then perfect CSI is replaced by estimated CSI in the decision metric. The expression for the moment generating function (MGF) of the received signal-to-noise ratio at the destination node is derived. Using the MGF expressions, we obtain the analytical symbol error rate for the proposed channel estimation based decoder for general OSTBCs with M-ary phase-shift keying modulation symbols.